What makes a good innovation partnership?

Can health systems incorporate disruption without introducing risk? In a word, no. But when startup innovators and health systems manage risk effectively, good partnerships emerge, and innovation thrives.

At the Chicago-Israel Health Tech Innovation Summit, several innovation communities came together to explore how health IT startups can successfully bring their inventions to market. I had several constructive conversations with entrepreneurs and shared insights into the complex world of doing business with integrated delivery networks.

It can be challenging for small startups to attract the attention of large health systems and then earn enough confidence to win the deal. The best way for them to begin is to listen well and engage in real conversations about the challenges the IDN is working to solve. Many of these innovators have great ideas and technologies, and when they understand the operational realities and context of a large health system, it can create advantages for both sides.

Unfortunately, many startups view these relationships as merely transactional, instead of as true partnerships. The entrepreneurs that show some level of investment or commitment to the solution’s performance—whether that means providing an expert onsite or tying their revenue to the IDN achieving certain outcomes—these are the businesses that will succeed.

Think through the whole workflow
Most innovators are selling a “killer app” for one component of a broader clinical workflow. Because IDNs have complex, interrelated clinical processes and workflows, startups need to think through how the solution may affect everything else. We cannot afford to disrupt the whole process, even if the application solves one problem in the most ingenious way.

For example, when a physician orders a medication, it is part of a must larger medication reconciliation process with interconnected workflows. There are considerations for nursing and pharmacy, regulatory requirements, as well as accurate tracking from ordering to dispensing to administering to billing, among others downstream—such as risk classifiers which create worklists to help care managers identify and coordinate care for higher-risk patients. A new app that addresses one piece within this complex set of clinical processes must account for the ways in which it may affect other aspects.

It’s also important to minimize the impact the new tools may have on clinician users. Our workforce contends with competing demands from its technology tools. To complete required tasks within their daily routine, they may already be interacting with a multitude of apps; it’s not appealing to consider launching yet another solution to do yet another subset of work.

IDNs must continuously improve to meet evolving organizational priorities and regulatory mandates. We are always updating our technology toolbox to meet a variety of demands, and at any given time, we have dozens of active projects in the works that will impact our clinicians. And we must do all of this without contributing to clinician burden, a growing industry concern.

Expect complex requirements and longer timelines
Having an open electronic health record (EHR) platform goes a long way to solving integration issues. And Northwell has that in our Allscripts EHR.

But even with an open EHR, IDNs have tremendous size and scope that create complex integration landscapes, which expand our testing requirements. We have myriad front-line systems, multiple layers of identification management and are managing literally thousands of interfaces across the enterprise. So, when a health IT startup pitches a new solution and is excited to tell us, “All we need is an ADT [HL7 Admission, Discharge, Transfer] interface to get started,” we still have a long way to go before we can successfully integrate this application.

Testing must address a wide range of scenarios. When thinking of a solution for the emergency department, for example, innovators must think beyond the expected scenarios of “patient is admitted” and “patient is discharged.” In one of our smaller hospitals, we need to test 23 separate scenarios representing how a patient might get registered just for an emergency department visit.

In addition to the complexity of our testing requirements, we have more intricate needs for project management. Startups often like to use the adaptive (agile) methodology for project management. But the large and interconnected nature of IDNs, specifically the interdependencies between subgroups within IT, requires a level or predictability and scheduling rigor that often cannot be achieved with adaptive methodologies. At each phase of any project, there are numerous stakeholders involved. It is common for major projects to pull them in different directions and cause collisions of priorities.

Address privacy and security concerns
However, one priority remains foundationally important: data security. The healthcare industry is a target for cyber criminals, and those attacks continue to increase in this sector.

According to the FBI Cyber Division, cyber threats against healthcare and medical devices are increasing, because of the potential for financial gain. It estimates that even a partial patient record is worth about $50 on the black market, compared to $1 value for a stolen Social Security or credit card number. For many IDNs, this valuation drives risk exposure into the tens of billions of dollars.

Large health systems must prioritize privacy and security. We conduct thorough vendor risk assessments, and entrepreneurs must demonstrate best practices to gain a seat at the table. Health systems must protect data and ensure that every innovation—no matter how small—is not a gateway into our networks for threat actors.

Get excited about “dirty data”
Clinical documentation is not used solely to document and direct clinical care; administrative and billing concerns often impact what and how things are documented. Other times, data is stale because of infrequent visits with the patient, and clinicians are unlikely to go back into patient records months or years later to update their notes on patients not actively being seen in their practice.

And in some cases, observations with dual documentations often show discordant values between physicians and nurses. This might happen when, for example, an elderly patient tells a physician that she no longer smokes, but the patient’s son informs a nurse that she has not broken the habit. Each clinician’s documentation will reflect the information that they received, leading to discordance within the record.

These are the realities of data within a large health system, often coming from multiple source electronic health record (EHR) systems. The quality of clinical data is less than ideal. Startups cannot presume pristine streams of information coming into their applications, but instead should be ready to embrace “dirty data” they are likely to find in IDNs.

Solutions must have clinician involvement and advocacy
The IT side of IDNs recognize that, unless clinicians are clamoring for the solution, it will be a tough road to adoption. Clinicians want to hear about benefits of a new solution from their peers—not from IT, and not from the entrepreneur trying to sell it.

At Northwell Health’s Innovation Center, I help identify the clinical stakeholders who can provide expertise in identifying the most valuable new technologies. Engaging them is critical in these early stages, and clinicians exhibit great passion and talent in a wide range of areas.

For example, collaborations through the Innovation Center are making some significant advances in monitoring, following the rapid adoption of sensors and wearables. We expect vast amounts of data from these devices, and now technology must address how to process and interpret this information for best results.

We’ve also explored UV light disinfection techniques, using drones to deliver materials into trauma situations as well as voice-based systems for patients and clinicians. These are just a few examples of the technologies we’re experimenting with at the Innovation Center, bringing together individuals who are thinking creatively about the toughest challenges facing today’s clinicians.

Each time a startup approaches an IDN, it’s an opportunity to improve healthcare. But, like bricks in a mosaic, it takes time to fit everything together. I hope startups don’t view IDNs as resistant to new ideas, but rather we’re making sure everything we layer on top of our EHR will do some good. The pieces must all contribute and work together to create something valuable—for the innovator, for us and, most importantly, for our patients.